> > >All endianness problems should be handled by ntoh... and
> > >hton... macros/functions. I have never checked whether everything is
> > >really done right, but if handled correctly, things like that don't
> > >cause problems.
> >
> > Is OpenSSL running on any big-endian machines?
Most OpenSSL algorithms are endian neutral (with exception for ripem
digest which is broken on big-endians, fix is on the way). Then there're
few that are endian "aware." "Awareness" is optional though and gets
enabled by feeding the compiler with -DL_ENDIAN or -DB_ENDIAN command
line option (well, not Mac compiler, but it's possible to predefine
macros in project preferences, right?). Enabling "awareness" (normally)
improves performance of algorithms in question.
> (Uh oh-- I always get
> > that wrong. I believe that x86 CPU's are little-endian, and PPC's are
> > big-endian.
It's easy:-) There're very few little-endians and a lot of big-endians.
"Few/lot" are in terms of variety, not amount of sales:-) "Few" are
Intel-based and Digital boxes (VAXes, Alphas and old MIPS-based
DECstations) and "a lot" are PPC, SPARC, MIPS, PA-RISC, 68k, whatever...
Yes, I know that PPC can run in little-endian mode, but I haven't seen
no little-endian OSes for PPC.
Speaking about OpenSSL for Macintosh. In March a guy wrote to
openssl-dev:
> From: CJ Holmes
> Subject: OpenSSL on the MacOS
> Date: Wed, 3 Mar 1999 23:56:36 -0500
>
> I had just finished porting SSLeay 0.9.0 to MacOS 8.x when I learned (quite
> by accident) that it had become OpenSSL. I was at once elated that the
> library would continue to evolve, and consternated that a 0.9.1 had been
> available for some time!
See http://www.mail-archive.com/openssl-dev@openssl.org/msg00483.html
for complete post. As far as I remember he was encouraged to post
changes. Has anyone heard from the guy again?
Andy.
______________________________________________________________________
OpenSSL Project http://www.openssl.org
Development Mailing List [EMAIL PROTECTED]
Automated List Manager [EMAIL PROTECTED]
